Table 1. The evaluation performance of the models in the pipeline.
Random Forests (RF) Run | RF+Stepwise | RF+Ridge regression | RF+Lasso | |||
---|---|---|---|---|---|---|
#Iterations | Number of trees (ntree) | % Variance all (p) SNPs | % Variance k SNPs | Model significance (r-squared & p-values) | Model significance (%Variance) | Model significance (r-squared & p-values) |
1 | 500 | -0.67 | 8.58 | 0.088 & 4.965e-09 | 14.7 | 0.096 & 1.83e-09 |
2 | 1000 | 0.43 | 13.7 | 0.11 & 4.771e-11 | 17.83 | 0.10 & 8.1e-10 |
3 | 2000 | 0.23 | 16.66 | 0.111 & 5.4e-11 | 18.77 | 0.138 & 7.108e-13 |
4 | 3000 | 0.51 | 17.94 | 0.13 & 1.49e-12 | 17.5 | 0.16 & 6.61e-15 |